Objective
The objective of the study was to predict the outcome of dichorionic (DC) twin pregnancies at 6-10 weeks’ gestation from intertwin discordance in crown-rump length (CRL), heart rate (HR), and gestational sac diameter (GSD).
Study Design
Intertwin discordance in CRL, HR, and GSD, was compared in 3 groups of DC twin pregnancies with 2 live embryos at 6-10 weeks. In one group, there were 2 live births (n = 174), in another 1 live birth (n = 24), and in a third group there were no live births (n = 33). Regression analysis was used to determine the predictors of outcome.
Results
Significant contributors to the prediction of 1 intrauterine death were discordance in CRL and GSD. The only predictor of 2 deaths was discordance in CRL. Detection rates of screening for single or double death, at 10% false-positive rate, were 79.2% and 27.3%, respectively.
Conclusion
In DC twins, intertwin discordance in CRL and GSD at 6-10 weeks can predict outcome.
In singleton pregnancies an early ultrasound examination before 10 weeks’ gestation through a series of measurements, including embryonic crown-rump length (CRL), embryonic heart rate (HR), and gestational sac diameter (GSD) can provide a useful prediction of subsequent outcome.
There is also some evidence that an early scan may predict outcome in twin pregnancies. One study examined a total of 77 monochorionic (MC) and dichorionic (DC) twin pregnancies at 7-9 weeks and again at 11-13 weeks. They reported that in the 9 cases with 1 live and 1 dead fetus at the second scan, compared with the 68 with 2 live fetuses, the intertwin discordance in CRL at the first scan was significantly higher. At a cutoff CRL discordance of 20%, the detection rate (DR) of 1 death was 77.8% at a false-positive rate (FPR) of 7.4%.
Similarly, another study investigated the value of an ultrasound scan before 8 weeks in DC twins with 2 live embryos in the prediction of subsequent death of 1 embryo/fetus compared with double-live birth. In 127 cases at 5-6 weeks, they measured GSD, and in 126 at 6-8 weeks, they measured CRL. For a discordance of 3 mm in GSD, the DR for a subsequent single loss was about 26% at FPR of 6.5%, and the respective values for discordance of 3 mm in CRL were 39% and 12.5%.
The aim of this study was to investigate further the potential value of an early ultrasound scan in DC twins in the prediction of subsequent pregnancy outcome.
Materials and Methods
In our hospital, there is an early pregnancy unit (EPU), which is freely accessible to pregnant women in our area. On arrival the demographic data and obstetric history are recorded in the EPU database and an ultrasound scan is carried out. Maternal characteristics recorded include age; racial origin (white, African, South Asian, East Asian, and mixed); method of conception (spontaneous or assisted conception requiring in vitro fertilization); cigarette smoking during pregnancy (yes or no); and parity (parous or nulliparous if no delivery beyond 23 weeks). The indications for attending the EPU are classified as vaginal bleeding, abdominal pain, anxiety because of previous miscarriage or ectopic pregnancy, and pregnancy dating. The objectives of the ultrasound scan at 6-10 weeks’ gestation include the diagnosis of an intrauterine or extrauterine pregnancy and, where appropriate, recording of the number of live or dead embryos and measurement of CRL, HR, and GSD. Ultrasound findings are recorded in the EPU database at the time of the scan.
The CRL was measured in a sagittal section of the embryo with care being taken to avoid inclusion of the yolk sac. The HR was calculated as beats per minute by the software of the ultrasound machine after measurement by electronic calipers of the distance between 2 heart waves on a frozen M-mode image. The GSD was calculated as the average of 3 perpendicular diameters with the callipers placed at the inner edges of the trophoblast.
Twin pregnancies were classified as DC or MC according to the presence of the lambda or T sign as reported previously. Gestational age was calculated from the CRL of the larger embryo.
We searched the EPU database to identify all women who attended the unit on at least one occasion when the scan demonstrated a DC twin pregnancy with 2 live embryos ( Figure 1 ). We then obtained details on pregnancy outcome of these patients from the computerized database of the maternity unit or the general practitioners of the women. On the basis of outcome, the pregnancies were divided into 3 groups: in 1 group, both embryos/fetuses subsequently died; in a second group, 1 embryo/fetus subsequently died and the other was live born; and in a third group, there was a subsequent live birth of 2 phenotypically normal neonates. In this study group, there were no cases of aneuploidies or major defects identified either prenatally or postnatally.
The study was examined by the Research Ethics Committee (REC) of the hospital, and it was decided that it did not require formal REC approval.
Statistical analysis
The 3 outcome groups (single death, double death, and double-live birth) were compared by Kruskal-Wallis test followed by a post hoc Dunn’s procedure for continuous variables and Fisher’s exact test or χ 2 test for categorical variables.
Logistic regression analysis was performed to determine the significant contributors to the prediction of 1 death or 2 deaths from maternal age; racial origin; method of conception; indication for attending the EPU; and intertwin discordance in CRL, HR and GSD, expressed as a percentage difference from the largest measurement. The performance of screening was determined by receiver-operating characteristic (ROC) curves.
The patient-specific risk for 1, 2, or any fetal deaths was calculated from the formula: odds/(1 + odds), where odds = e Y . The Y for the relation between 1, 2, or any fetal deaths and discordance in CRL (percentage) was derived by univariate logistic regression analysis.
The statistical software packages PASW statistics 18.0 (SPSS Inc, Chicago, IL) and XLSTAT 2009 (Addinsoft, New York, NY) were used.
Results
The data search identified 245 DC twin pregnancies that were examined in the EPU between February 2006 and May 2009. Excluded from further analysis were 14 subjects that were lost to follow-up. The maternal characteristics and indications for attending the EPU in the remaining 231 subjects are summarized in Table 1 .
